Progress in biomedical research and its translation into clinical practice require the integration of data across multiple scales (molecules, cells, organisms), organism types, and fields of research. The need for data integration is especially acute in infectious disease research where organisms interact on all scales, and these interactions result in the emergence of processes and structures specific to these interactions. True data integration, the ability to jointly interpret and analyze data of heterogeneous types, depends on the ability to link data to information about the biological entities to which the data refer. In the face of rapidly growing volumes of data and information, it is imperative that this link from data to information be computable. Automated processing of the links between data and information requires that they be expressed using a common, formalized system for knowledge representation. Efforts at knowledge representation in biology have focused on either ontology development or pathway representation. While the value of both is unquestionable, neither fully supports the data and information integration needs of infectious disease research. We propose an ontology-based approach to pathway representation that extends ontologies beyond single taxonomies and pathway representations to all levels of granularity, thereby allowing the representation of complex biological systems. Our approach builds upon existing ontologies and pathway representations but is grounded in formal ontological and logical principles. Our overall goal is to test empirically the degree to which the ontology-based representation can improve data interpretation and analysis for translational medicine. We will take as our case study Staphylococcus aureus infection, utilizing the invaluable data resources of the Duke Staphylococcus aureus Bacteremia Group. We will achieve our goal through the following three specific aims: 1. Create an ontology-based representation of host-pathogen interactions, focusing on Staphylococcus aureus bacteremia. 2. Empirically test the ability of the ontology-based representation created in Aim 1 to improve data analysis and interpretation by using the representation to predict disease genes associated with Staphylococcus aureus bacteremia. 3. Empirically test the impact of the ontology-based representation created in Aim 1 on understanding of Staphylococcus aureus pathogenesis, on identification of novel therapeutic targets, and on improvement to patient management by testing experimentally the disease gene predictions made under Aim 2. The anticipated outcomes are: an ontology-based method for the representation of complex biological systems and an ontology of host-pathogen interactions, both subjected to tests designed to demonstrate their utility to clinical and translational research; an improved understanding of the immune response to bacterial pathogens; and the identification of genes associated with Staphylococcus aureus bacteremia that can be used to develop novel diagnostics and therapeutics.The resources developed under this proposal will directly improve data integration, retrieval and analysis, will support cross-disciplinary collaborations within infectious disease research, and will provide a foundation from which to develop similar resources for other areas in biomedicine, thus significantly impacting biomedical research and translational medicine. [unreadable] [unreadable] [unreadable]